Staphylococcus spp., Streptococcus spp. (Groups C and G), Pasteurella spp., Corynebacterium spp., Actinomyces spp., hemolytic Escherichia coli, and Aeromonas spp. cause abscesses and localized pulmonary, hepatic, uterine, vulvar, epidermal, mammary, and oral infections.
Penetrating injuries, such as bite wounds inflicted during mating and oral puncture wounds from ingested bones and foreign bodies, can result in abscess formation. In most cases, infection is locally sequestered, resulting in few systemic clinical signs. Sound management practices can reduce environmental and feed-associated sharp object exposure and limit contact time between males and females during breeding, thus lowering abscess incidence.
Subcutaneous abscesses are detected as fluctuant swellings that may be accompanied by draining tracts (Fig. 21.1). These must be distinguished from cutaneous myiasis. Abscesses or necrosis of internal organs can be noted clinically by organ-dependent signs, lethargy, and pyrexia [1].
Aspirating, draining, or biopsying the affected swelling will aid in differentiating inflammation from neoplasia or parasitic infestation. Culture and antibiotic sensitivity tests of isolated microorganisms will assist both in diagnosis and treatment.
Localized abscess drainage and topical antiseptic application are usually effective. Systemic antibiotics should be initiated if drainage and localized treatment do not eliminate infection and ameliorate clinical signs.
This disease, which is rarely reported in ferrets, is often called lumpy jaw. In cattle, the disease is caused by Actinomyces bovis, while in other animals, including humans, Actinomyces israelii is the responsible agent [2–5]. This condition is also documented with some frequency in cats with feline immunodeficiency syndrome or feline leukemia virus infection, and in HIV-positive humans. Other Actinomyces species have also been associated with disease and may reflect normal oral flora invading compromised tissue.
Disease predilection for the cervical area reflects organism entry through damaged oral mucosa. Feeding ferrets carcasses with bones may increase the likelihood that actinomycosis, if present in oral flora, could infect animals through cuts and abrasions. The bacterium may also be swallowed or inhaled. The source of clinical disease is considered endogenous, although animal bites may also transmit it. Given the few documented ferret cases, Actinomyces susceptibility is probably low. However, immunosuppressed ferrets may be at an increased infection risk.
Cervical masses with sinus tracts containing thick, greenish-yellow, purulent material were recorded in one actinomycosis case. The masses enlarged, and the animal had difficulty breathing. On histologic study, cervical tissue contained abscesses with characteristic Actinomyces organisms located throughout (6). In a second case, the animal died without apparent clinical signs, and firm nodules were found beneath the visceral pleura. The posterior mediastinal lymph nodes were twice as large as normal, with abscesses containing greenish granules present on the cut surface. Histologically, the abscesses contained typical Actinomyces colonies [6]. Three other cases of suggestive neck swelling have been diagnosed in research ferrets [2]. We have also observed a subcutaneous Actinomyces granuloma in a ferret with multicentric lymphoma [7].
Actinomyces isolation from aspirated exudates is required for definitive diagnosis. All species of these bacteria are gram-positive, non-acid-fast, anaerobic to microaerophilic organisms. The disease alternatively can be diagnosed through histologic demonstration of organisms in specially stained tissue sections. While these bacteria are rod-shaped, they can appear similar to fungal hyphae due to their formation of branched networks. Actinomyces-associated cervical masses must be distinguished from other causes of cervical swelling, including neoplasia, cervical granulomatous masses, and sinuses due to Hypoderma bovislarval myiasis, and staphylococcal or streptococcal cervical cellulitis secondary to dental disease-associated mandibular osteomyelitis [6]. Due to the radial growth of the organism, colonies form in the affected tissue, surrounded by an inflammatory response, resulting in firm, yellowish granules called sulfur granules. These granules also may be present in purulent draining material from the diseased tissue.
Sustained treatment with high-dose penicillin or tetracyclines is effective in treating human disease. Antibiotic treatment should be supplemented with surgical abscess drainage, local debridement, and topical antibiotic and antiseptic application. Fluid and nutritional support should also be considered in light of potential oral pain, which could be mitigated through the administration of nonsteroidal anti-inflammatory drugs (NSAIDs).
Various organisms have been implicated in primary or secondary bacterial pneumonias. Streptococcus zooepidemicus as well as Group C and Group G streptococci were cited as primary pneumonia causes and were isolated from the lungs secondary to influenza infection [2,8]. Streptococcus pneumoniae, which has been identified as a cause of mink pleuritis and pneumonia, is also a potential cause of ferret bacterial pneumonia [9]. Gram-negative bacteria, such as E. coli, Klebsiella pneumoniae, Pseudomonas aeruginosa, and Pseudomonas luteola have also been isolated from suppurative lung lesions [10–12]. Bordetella bronchiseptica was isolated from diseased neonatal ferret lungs [13]. Listeria monocytogenes was also isolated from consolidated inflammatory lung tissue and pleural effusion of a ferret with Cushing's syndrome and cardiomyopathy [14].
Nasal discharge, dyspnea, increased abdominal respiration, fever, lassitude, and anorexia may be observed. In certain cases, fulminant pneumonia with sepsis causes acute death without preexisting clinical signs. In one such case, peracute respiratory disease of presumed bacterial etiology (possibly B. bronchiseptica), was observed in 7- to 10-day-old kits. The kits presented with dyspnea, open-mouthed breathing, and serous to suppurative nasal discharge. Without antibiotic treatment, they died within 24 hours [13].
Bacterial pneumonia is usually characterized by a mixed inflammatory cell process, either primarily bronchial or diffusely affecting the lung. Streptococcus Group C infection is characterized by purulent bronchial material and raised, variably sized (up to several millimeters in diameter) yellow to white patches on the dorsal and marginal lung surfaces [2]. Several ferret studies emphasize the importance of influenza-bacterial synergism in clinical outcome. One early study indicated that ferrets have a more severe infection when challenged with Streptococcus Group C and influenza when compared with either disease alone [15]. Similar results have been obtained with S. pneumoniae [16]. Another study indicated that ferrets were susceptible to aerosol Streptococcus Group C challenge only if first infected with influenza A virus [17]. Others have shown that influenza A virus infection significantly enhances Staphylococcus aureus adherence to desquamating anterior turbinate respiratory epithelial cells. In this study, the amount of staphylococci recovered (60–90 minutes after dosing) from virus-infected ferrets was significantly higher when compared with controls, particularly 2–5 days after viral challenge [18].
Another report demonstrated that staphylococcal α and γ toxins, along with endotoxin and diphtheria toxins, were lethal to 5-day-old ferrets [19]. In those infected at 1 day of age, this toxicity was enhanced 3-fold for staphylococcal γ toxin, 14-fold for staphylococcal α toxin, 84-fold for endotoxin, and 219-fold for diphtheria toxin. Pathologic assessment identified upper respiratory tract inflammation, lung edema and collapse, and early bronchopneumonia only in ferrets treated with both toxin and influenza virus, but not in those treated with virus or toxin alone [19].
Definitive diagnosis is based on clinical signs and culture from clinical specimens, such as tracheal exudate, or affected lung tissue. Animals stressed from concurrent infectious disease such as influenza, debilitated from chronic illnesses such as cardiomyopathy, or irnmunosuppressed from therapy, metabolic disease, or surgery may be more susceptible to bacterial pneumonias. Differential diagnoses include Dirofilaria immitis-associated pleural effusion, dilated cardiomyopathy, pulmonary mycosis (e.g., Pneumocystis carinii in immunosuppressed ferrets), primary or secondary neoplasia, malignant hyperthermia, as well as asthma or allergies.
If bacterial isolation is successful, in vitro antibiotic susceptibility tests should be performed. Generally, penicillin or synthetic penicillins are effective against gram-positive bacteria, such as streptococcal organisms. Trimethoprim and sulfadiazine (30 mg/kg) is effective in treating susceptible gram-negative bacteria. One large commercial ferret supplier previously vaccinated kits with B. bronchiseptica bacteria and had claimed its use curtails weanling ferret bacterial pneumonia. Its use, however, at their breeding site is no longer applied in their preventative disease program.
Botulism is a neuroparalytic disease usually caused by consumption of foods containing toxin(s) produced by anaerobic spore-forming bacteria [20].
Botulism in fur production animals has been extensively documented. Outbreaks in mink associated with type C toxin ingestion have resulted in 90% mortality rates [21]. Similar high mortality rates were documented in Clostridium botulinum type A-infected mink [22]. Ferret deaths attributed to C. botulinum toxin type C-contaminated wild bird carcass ingestion have been recorded in England [23,24]. Experimentally, ferrets are susceptible to C. botulinum types A and B infection, highly susceptible to type C, but refractory to type E [25,26]. Disease depends in part on soil-borne clostridial spores, and C. botulinum type B spores are far more prevalent in the midwestern United States than other clostridial types. Disease is typically attributed to feed contamination. As a result, ferrets should be fed only as much as they can immediately consume, and excess food should be removed to prevent hoarding. Food should be fully cooked, refrigerated where appropriate, and dry foods should contain a preservative. However, some C. botulinum strains, especially type E, can grow in cold or acidified environments, though type C strains, which appear to be most common in ferrets, require at minimum 10°C and pH 5.1 [27]. Annual toxoid vaccinations are recommended for commercially produced ferrets, and these appear to provide effective disease protection [28]. Due the limited availability of several ferret-specific distemper vaccines, some ferret breeders have utilized Distox-Plus for distemper protection [29]. Labeled for use in mink, this product contains modified live and inactivated mink distemper virus, C. botulinum type C toxoid, and inactivated P. aeruginosa (four serotypes). This product could be used to prevent disease in production settings or where disease risk is high.
Ferrets orally inoculated with toxin types A and B develop signs within 12–72 hours, and death ensues in 1–7 days [25]. Clinical signs initiate with blepharospasm, photophobia, lethargy, and urinary incontinence. Rapid weight loss, ataxia, and ascending paralysis are also noted as the disease progresses (Fig. 21.2). Death typically results from respiratory failure resulting from paralysis of the intercostal muscles and diaphragm. The disease course is variable and to a large degree independent of toxin exposure level. Clinical recovery seldom occurs without treatment. The female ferret appears to be more susceptible to toxin-mediated effects than the male; this may be due to the female's smaller size. Hepatic, splenic, and renal congestion with splenic subcapsular hemorrhage is common. Focal cerebellar hemorrhage is also consistently noted. Histologically, there is significant splenic lymphoid depletion with central follicular necrosis, follicular mononuclear cell hyperplasia, disruption of the normal reticular pattern, and a marked increase in multinucleated giant cells. This also occurs to a lesser extent in the lymph nodes. Hepatic congestion with capillary and venous distension as well as occasional hemorrhage is also occasionally noted.
Diagnosis is based on the clinical history of ingestion of spoiled or improperly handled food, followed by the development of central nervous system (CNS) symptoms and rapid weight loss. Gross lesions and histologic changes in affected ferrets are similar with ingestion of toxins A and E, and severity depends on the length of illness [25]. Definitive diagnosis is based on demonstrating the presence of botulinum toxin in a ferret's blood, serum, gastrointestinal tract, or in contaminated food. Classically, mouse protection assays utilizing monoclonal antibodies were used to detect and type toxins present in patient samples, though PCR-based assays have recently been described [30,31]. However, neither of these assays addresses C. botulinum type C detection, which may be more common in ferrets. Anaerobic culture may also be utilized and toxin detection assays performed on culture supernatant. Botulisum should be differentiated from other causes of neurologic signs in the ferret, particularly canine distemper and rabies. However, acute disease onset and potentially rapid deterioration merit empirical treatment in suspect cases.
Once botulinum poisoning is suspected, aggressive fluid and nutritional support should be instituted. Prokinetic drugs can be given to facilitate the organism's clearance from the gastrointestinal tract, and diuretics can be given if renal parameters merit. Penicillin can be given to clear the organism, though antibiotic use is controversial due to the potential for sudden toxin liberation associated with bacterial death and their detrimental effects on normal intestinal flora [32]. Antitoxins are not routinely available.
Campylobacteriosis is a diarrheal disease caused by the gram-negative microaerophilic bacterium Campylobacter jejuni and less frequently by Campylobacter coli. The genus Campylobacter currently consists of 17 species, 14 of which have been isolated from humans [33]. This is an important cause of human diarrhea and has been linked to diarrhea in ferrets, mink, dogs, cats, and nonhuman primates [34].
Historically, Campylobacter-like organisms were thought to cause proliferative colitis in ferrets. However, it is now known that Lawsonia intracellularis causes this condition and that neither C. jejuni nor C. coli plays a role in its pathogenesis. C. jejuni is now included with Salmonella spp. and Shigella spp. as a leading cause of human diarrhea. Several reports have linked human disease to household pets, particularly dogs and cats recently obtained from animal shelters or pounds [35]. In C. jejuni prevalence studies, most positive ferrets were asymptomatic carriers. In one report, nondiarrheic feces from two commercial sources were screened, and 8 of 10 and 43 of 73 cultures taken from the two vendors were C. jejuni-positive [36]. It is not known how long ferrets naturally infected with C. jejuni shed the organism in their feces, but experimentally infected ferrets shed for up to 16 weeks after inoculation [37]. Following experimental inoculation, 108–109 colony-forming units (CFUs) are shed per gram of feces through day 9 post infection. A sharp shedding peak is noted at day 3, where 1010 CFUs per gram of feces are seen [38]. The high prevalence and long-term shedding of C. jejuni presents an occupational hazard to laboratory animal personnel and households with pet ferrets [39].
Foodborne and waterborne transmission appear to be the principal avenues of ferret infection, with the ingestion of contaminated meat products, such as poultry and unpasteurized milk, being particularly important. Mink kit diarrhea epizootics caused by C. jejuni and C. coli have been described and were likely associated with feeding of uncooked chicken offal [40]. Nosocomial infections are also possible, as is exposure to other pets (dogs, cats, hamsters, and birds) and to farm animals that may shed the organism [41,42]. Such transmission events are of particular concern, given the disease's subclinical nature and potential for long-term shedding and subsequent zoonotic or cross-species transmission.
Given the high prevalence of natural disease, it is not surprising that ferrets have been used to study C. jejuni-induced diarrhea and its relationship to acquired and maternal immunity [41,43,44]. In this model, self-limiting diarrhea is consistently produced. As shown in the primate model of Russell et al. [45] and the removable intestinal tie adult rabbit diarrhea (RITARD) model [46], acquired immunity provides at least partial infection resistance, manifested mainly by shortened C. jejuni colonization periods and serum antibody development. While antibodies do not prevent C. jejuni colonization, they do provide protection from diarrheal illness [37,43]. Cross-protection between C. jejuni strains has also been noted, and one study examining three different strains noted protection of 67–100% in all cases [47]. In the same report, formalin-killed C. jejuni preparations administered intragastrically provided similar protection when 4 doses were given 48 hours apart.
Clinical diarrhea has been experimentally induced in ferrets inoculated orally with C. jejuni [44]. As in cases of Campylobacter infection in dogs and cats, ferrets can be asymptomatic carriers. Clinical diarrhea appears to occur more frequently in animals younger than 6 months of age. In addition, animals may be more susceptible to clinical infection if stressed by experimental regimens, hospitalization, concurrent disease, pregnancy, shipment, or surgery. The pathogenic mechanisms responsible for the clinical alterations and laboratory changes are poorly understood. In our experience, the typical presenting clinical signs include mucus-laden, watery, or bile-streaked diarrhea (with or without blood and leukocytes) of several days duration and partial anorexia [44,48]. Ferrets experimentally infected with C. jejuni consistently produce diarrhea within 24 hours, and 80% of infected animals have grossly apparent blood in their stool for up to 3 days [38]. In natural and experimental C. jejuni-associated colitis in mink kits, yellow-white mucoid diarrhea, often blood-tinged, has been noted [40]. Fever, rectal prolapses, and occasional leukocytosis are also seen. Adults are not affected.
In ferrets, diarrhea can persist for more than 4 weeks or may be intermittent [40]. Histopathologic changes in ferrets experimentally infected with C. jejuni consist of colonic goblet and epithelial cell loss with neutrophilic inflammation and concurrent neutropenia, particularly in the first several days following infection [37,44]. However, concurrent cryptosporidiosis, although reported to be nonpathogenic, makes lesion interpretation difficult in these cases [49]. Bacteria are noted on the colonic luminal surface, as well as within and between enterocytes. On electron microscopy, bacteria are seen attached to the epithelium with loss of associated microvilli. Intracellular bacteria are free within the cytoplasm. Following peracute disease, histopathologic changes become much less pronounced with epithelial coverage recovering by day 6 post infection [38]. Despite its reputation for colonic tropism, C. jejuni can invariably be cultured from both the small and large intestine, as well as from the livers in 78% of cases. The mechanism of hepatic translocation is not known, but presumably bacteria either travel retrograde through the biliary tree or are transported there by macrophages. Studies with the RITARD model have also demonstrated that the M cell found in the epithelial layer over intestinal Peyer's patches is a probable route for systemic spread of C. jejuni [50].
In addition to diarrheal disease, mink abortion caused by C. jejuni has been reported [51]. Experimental infections have proven that C. jejuni can also induce abortions in ferrets. In one study, four pregnant mink and seven pregnant ferrets, including five with previous exposure and specific C. jejuni antibodies, were injected intravenously with 108–1010 CFUs of C. jejuni [52]. All 11 animals aborted 1–16 days after infection, with results ranging from fetal reabsorption to delivery of dead or premature living kits. In every case, uterine contents (placenta, uterine fluid, and/or kits) were C. jejuni culture-positive. In a second experiment, three pregnant mink and nine pregnant ferrets, including four with previous exposure and antibodies, were fed 109–1011 C. jejuni CFUs. Two of the mink aborted and kits from all three mothers were culture-positive [52]. Seven of the nine ferrets aborted, with two having C. jejuni-positive uterine contents. None of the 28 uninfected ferret control pregnancies ended in abortion. Histologically, severe placentitis was evident, and this appears a more likely cause of Campylobacter-induced abortion than direct pathogenic effects on infected fetuses. These results suggest that mink or ferret C. jejuni infection during pregnancy poses the risk of reproductive failure, even for females with previously acquired immunity.
Campylobacter infection must be distinguished from other causes of diarrhea, including E. coli, Salmonella spp., and L. intracellularis. Because of its slow growth and the need for microaerobic conditions, standard methods used for culture require either filtration or selective media that incorporate various antibiotics to suppress competing fecal microflora. Cultures are incubated for 48–72 hours at 37°C and 42°C in an atmosphere containing 5–10% carbon dioxide and an equal amount of oxygen. C. jejuni colonies are round, raised, translucent, and sometimes mucoid. The organism can be identified by a series of biochemical tests readily available in diagnostic laboratories. Hippurate hydrolysis distinguishes C. jejuni from C. coli (Table 21.1). Biochemical and molecular assays should also be used to ensure that Helicobacter mustelae is not incorrectly diagnosed as C. jejuni, given their morphologic and biochemical similarities. Such confusion is possible due to the relative ease in culturing H. mustelae from ferret feces [53].
Table 21.1. Bacteriologic Properties of Campylobacter and Helicobacter spp.a
While restriction fragment length polymorphism (RFLP) has traditionally been the gold standard for Campylobacter species discrimination, a multiplex polymerase chain reaction (PCR) assay has also been described. This technique was utilized to confirm C. jejuni identification in mink and silver fox endometritis cases and C. coli in a pig abortion case [54]. Thus, molecular analysis presents a promising new tool for confirmatory testing or in cases of ambiguous hippurate hydrolysis results. Campylobacter culture requires specialized equipment and precise technique, so the refinement of PCR-based diagnostic approaches presents an opportunity to streamline diagnosis and facilitate research. Molecular assays, particularly 16S rRNA sequencing, have resulted in the identification of several novel Campylobacter species [55].
Two widely adopted serotyping schemes used to identify C. jejuni and C. coli are the passive hemagglutination assay and the slide agglutination assay. The slide agglutination assay [56] uses heat-labile flagellar antigens (Lior serotype), whereas the passive hemagglutination assay [57] uses heat-stable antigens (Penner serotype), primarily lipopolysaccharide (LPS). Similar serotypes are recognized in both animals and humans [56,57]. Serotypes commonly isolated from humans have also been isolated from ferrets [58]. Although C. jejuni appears to be prevalent among ferrets, it is not known whether all ferret serotypes cause diarrhea in humans. Increased awareness of the zoonotic risks presented by ferrets, dogs, cats, and other domestic and laboratory animals, coupled with the utilization of reliable serotyping systems, will allow a more complete understanding of the animal host's role in human Campylobacter epidemiology [48].
In humans, therapy for Campylobacter enteritis is primarily supportive, consisting primarily of intravenous fluid administration to replace ongoing losses. While infections are typically self-limiting without antimicrobial therapy, erythromycin has been shown to clear C. jejuni from the feces of acute enteritis patients [59] and convalescent carriers [60] within 48 hours. Thus, antibiotics shorten the typical disease course, but only when given shortly following the onset of clinical signs. Because most patients do not seek medical treatment until they have been diarrheic for several days, antibiotics are typically not utilized. However, in large-scale ferret breeding operations, antibiotic treatment might be useful to reduce colony-wide ferret morbidity and mortality during epizootics once a definitive bacteriological diagnosis is established.
In vitro sensitivity studies indicate that C. jejuni is routinely sensitive to erythromycin, aminoglycosides, tetracycline, chloramphenicol, furazolidone, and clindamycin. Erythromycin is considered the drug of choice in C. jejuni-infected humans based on susceptibility, ease of administration, and the lack of serious drug toxicity [61]. Ampicillin, metronidazole, and sulfa combinations are less therapeutically effective. Several antibiotics, including penicillin, polymixin B, cephalosporins, trimethoprim, and vancomycin are ineffective [62–64]. Many Campylobacter strains produce a β-lactamase [65], which accounts for their penicillin resistance.
A therapeutic trial was undertaken in an attempt to eliminate C. jejuni from 16 carrier ferrets. Despite their uniform in vitro sensitivity to erythromycin, oral erythromycin therapy did not eliminate C. jejuni [48]. One possible reason is that the dose used was based on that recommended for severely infected children [59]. Consequently, interspecies differences in pharmacokinetics may account for the therapeutic failure. The lack of significant pretherapy and posttherapy antibiotic mean inhibitory concentration differences indicates that resistance did not contribute to therapeutic failure [48]. Reinfection with C. jejuni could also have occurred. One day after erythromycin treatment, 6 of 16 ferrets tested negative for infection. Subsequent intestinal recolonization of five of these six ferrets may have been facilitated by environmental contamination. Unfortunately, available microbiologic techniques and standard laboratory ferret housing conditions preclude a clear-cut distinction between a chronic carrier (with lack of therapeutic response) and reinfection [48].
However, erythromycin (220 g/ton of feed) successfully controlled diarrhea and colitis associated with two Campylobacter outbreaks in weaning mink kits. Supportive treatment included warming and hydrating the animals and frequent bedding changing to keep the kits dry [40]. It is therefore recommended to treat symptomatic ferrets with C. jejuni present in diarrheic feces. Re-culture of ferrets after treatment is recommended, and precautions should be taken to minimize zoonotic transmission as well as spread to other ferrets, household pets, or other laboratory animals.
In 1938, investigators isolated a chlamydial agent from ferrets inoculated with throat washings from influenza-infected humans [66]. In these cases, chlamydia induced pneumonitis in intranasally challenged ferrets. It also produced meningoencephalitis and pneumonitis in intraperitoneally or intracerebrally inoculated mice.
On necropsy, ferret lung lobes were firm, plum-colored, and edematous. Histologically, bronchiolar epithelia were hyperplastic with luminal exudate. The alveolar walls were edematous and densely infiltrated with large mononuclear cells. The alveolar spaces were distended, with the cellular exudate primarily composed of mononuclear cells with occasional polymorphonuclear cells [66]. Subsequent to these initial findings, the ferret chlamydial agent was used in numerous experimental studies [67]. It remains controversial, however, whether this agent originated from ferrets or humans. The strain is available for study from the American Type Culture Collection. Chlamydia spp. has been isolated from diarrheic ferrets, but its importance in diarrheal disease is unknown [68].
A C. perfringens type A gastroenteritis epizootic occurred in weanling black-footed ferrets maintained in a zoological resource center [69]. Although C. perfringens type A enterotoxemia has not been studied extensively, it is probably closely analogous to C. perfringens type D-associated disease. This agent also causes rabbit enterotoxemia, and ferret disease likewise displays a peracute onset as well as an association with diet change.
Because the organism is ubiquitous and its toxin weakly immunogenic, disease prevention is problematic. Dietary management is critically important in disease control. In the black-footed ferret outbreak, weanlings were fed a ration consisting of liver, rabbit meat, and mink chow pellets three times daily. It is likely that overeating or abrupt dietary change precipitated rapid bacterial proliferation and necrotizing toxin production. When a twice-daily feeding frequency was utilized, no more incidents of bloat syndrome were noted [69].
Affected ferrets die peracutely with marked abdominal distension, gastric bloat, and dyspnea. On necropsy, they display markedly distended, thin-walled gastrointestinal tracts containing excessive gas and moderate amounts of brown, semiliquid ingesta. Abundant gram-positive bacilli measuring 1 × 3–6 μm have been noted in smears of gastric and intestinal contents along with other mixed bacteria, epithelial cells, neutrophils, and erythrocytes. Histologically, disease is characterized by diffuse superficial to full-thickness mucosal necrosis with gastrointestinal epithelial sloughing. Gram-positive bacilli line the denuded mucosal surfaces and extend into the gastric glands and intestinal crypts [69]. Lymphoid necrosis of the lymph nodes, spleen, and thymus, attributed to physiologic stress, was also noted. Dilation of central hepatic sinusoids with centralobular hepatocellular dissociation and multifocal portal neutrophil aggregates were considered secondary to shock, toxemia, or bacterial emboli.
Acute bloat accompanied by a history of overeating or diet change is highly suggestive of C. perfringens enterotoxemia. Because of the potential for acute death, symptomatic animals should be empirically treated. C. perfringens can be isolated from gastric and small intestinal contents through anaerobic culture or suggestive organisms identified on gastric smears. As with C. botulinum, mouse protection assays have been used to detect and type toxin production by these bacteria [70], though PCR-based detection systems have also been described [71,72].
Ferrets presenting with severe gastric bloat and dyspnea should be handled as medical emergencies. Gastric decompression should be undertaken immediately either through orogastric intubation or, if unsuccessful, gastric trocharization. Radiography can be utilized to diagnose bloat, rule out the presence of a gastric foreign body, and confirm successful decompression. Serial radiography may be indicated if the condition fails to resolve after initial decompression. Intravenous shock fluid therapy should also be instituted, and an NSAID such as flunixin megalumine (1 mg/kg) should be used for up to 3 days to minimize the prostaglandin-mediated hypotension noted in endotoxemia. Antibiotics are likely of little value as they may contribute to dysbiosis and potentiate enterotoxemia due to abrupt bacterial death and toxin liberation. Oral administration of cholestyramine, an ion exchange resin capable of binding bacterial toxins, might be of value, given its success in treating rabbit enterotoxemia. Despite gastric trocharization and supportive therapy, ferrets typically die within 2 hours of clinical presentation [69].
E. coli has recently gained recognition as an important ferret pathogen. These bacteria possess several virulence factors important in other species, and thus, cross-species transmission appears possible. Additionally, ferrets have been used as an experimental E. coli infection model.
The earliest ferret E. coli case report described gangrenous mastitis in eight ferrets [10]. While coliform mastitis may occur (see Chapter 15, “Genitourinary Diseases”), E. coli is now also considered an important cause of ferret diarrhea and septicemia. This was first reported in zoo-housed, black-footed ferrets [73]. Five adults and three kits died suddenly, either without warning or following a 24-hour period of anorexia and loose mucoid feces. E. coli was cultured from rectal swabs and necropsy tissues. Seven of the eight isolates obtained were later shown to be heat-stable toxin (STa and STb)-positive enterotoxigenic E. coli (ETEC). One isolate tested ST-negative, but was cytotoxic necrotizing factor 1 (cnf1)-positive. All isolates were nonmotile and did not react with standard O-antigen antiserum. Black-footed ferrets in this cohort were fed a mix of mink chow, raw rabbit meat, beef liver powder, lard, and blood meal. ETEC was also isolated from food that was retrieved from an affected animal's cage. When the ration batch was replaced, no new cases were noted. These animals were housed separately from one another, so foodborne transmission was considered likely, but was never definitively confirmed. E. coli-associated septicemia was diagnosed in this cohort, as ETEC was cultured from the liver, kidney, or spleen of five of these black-footed ferrets [73]. In another report, septicemia was implicated in two neonatal black-footed ferret deaths, accounting for 3% of that institution's records on neonatal mortality [74]. Septicemia could explain the acute mortality seen.
Another study examined the virulence-related properties of 40 E. coli isolates from septicemic mink [75]. Twelve O serotypes were identified, the most common of which were O8 (12 isolates) and O6 (5 isolates). Because these and 18 other avian serotypes were identified, it was speculated that they were acquired through chicken offal and uncooked egg-based feeds. PCR-based molecular testing revealed that 48% of isolates were cnf1-positive and 8% were eae-positive [75]. Diverse patterns of antibiotic resistance were noted and may be prophylactic use-related. Resistance to tetracycline (83% of isolates), sulfamethoxazole (63%), streptomycin (60%), ampicillin (38%), and kanamycin (28%) were most commonly encountered. Multiple resistances were common, as 50% of isolates were resistant to four or more antibiotics. Conjugation experiments demonstrated R plasmid transmission in at least 27% of isolates. All isolates were susceptible to amikacin, ciprofloxacin, ceftiofur, ceftriaxone, cefoxitin, amoxicillin with clavulanic acid, and imipenem [75].
Other studies have also examined antibiotic resistance in farmed mink. One study compared resistance patterns in isolates from apparently healthy and diarrheic animals from different farms [76]. Tetracycline (22%), amoxicillin (23%), sulfamethoxazole (30%), spectinomycin (22%), and trimethoprim (10%) resistance was encountered, though farm-to-farm variability was high. All isolates were colistin, gentamicin, neomycin, and enrofloxacin susceptible. Hemolytic isolates, of which approximately equal numbers were obtained from diarrheic and healthy mink, were approximately twice as likely to display resistance. Another study similarly examined 1093 mink E. coli isolates cultured from both intestinal and extraintestinal (lung, liver, spleen, or urogenital) samples [77]. Fecal and intestinal samples most commonly displayed resistance to ampicillin (39%), streptomycin (26%), tetracycline (24%), sulfamethoxazole (24%), spectinomycin (11%), and trimethoprim (10%). These were also the most commonly encountered resistance phenotypes in extraintestinal E. coli isolates, though some site-specific variation was encountered. Very low resistance frequencies (less than 5%) were noted to fluoroquinolones, gentamicin, florfenicol, amoxicillin with clavulanic acid, ceftiofur, chloramphenicol, colistin, nalidixic acid, and apramycin.
cnf1-Positive E. coli have been identified in septicemic black-footed ferrets and mink [73,75]. This protein toxin is associated with human necrotoxigenic E. coli (NTEC) implicated in extraintestinal infections, including meningitis and approximately 80% of uncomplicated human urinary tract infections [78]. Ferret-origin E. coli isolates from rectal, urine, vaginal, milk, brain, and blood cultures produce functional CNF1, directly implicating this toxin in disease [79]. While cnf1 warrants further study in this context, its association with ferret and mink intestinal and extraintestinal disease merits its inclusion in virulence factor-based molecular screens.
While never definitively reported, ferret E. coli isolates have zoonotic potential. Ferret isolates share numerous virulence factors with human- and animal-origin pathogenic E. coli, such as eae, cnf1, and enterotoxins. Ferrets have also been experimentally infected with O157:H7 and other shiga toxin-producing E. coli (STEC) isolates. Infected ferrets maintained a local gastrointestinal infection and in 23% of cases developed evidence of renal disease, frequently including glomerular capillary fibrin deposition and hematuria. Glomerular damage and hematuria without bacteremia are consistent with the features of human hemolytic uremic syndrome [80]. In addition to this infection model, experimental pig and rabbit infections with bovine- and human-origin isolates further support that E. coli transmission is not host-dependent [81]. Evidence also suggests that isolates from animals and humans share a high degree of clonality [82], reinforcing the importance of animals as pathogenic E. coli reservoirs [83]. Human handlers should approach diarrheic ferrets with caution and utilize personal protective equipment to minimize transmission risks. Children should not be permitted contact with diarrheic animals due to their E. coli-induced hemolytic uremic syndrome susceptibility.
Ferrets infected with pathogenic E. coli will most likely present with diarrhea. Its duration, severity, consistency, and the presence/degree of hematochezia are strain-dependent. Diarrhea may assume a rapid course or could be several weeks in duration. Fever, dehydration, anorexia, and abdominal pain may also be present. Ferrets with colisepticemia present in shock or are found dead following 12–24 hours of anorexia or lose mucoid stool.
Upon necropsy, redness of the gastrointestinal mucosa may be noted. Also, intestinal contents may be bloody, mucoid, or watery in a strain-dependent manner. Histopathologically, ferrets display enterocolitis with large numbers of rod-shaped bacteria associated with intestinal villi [73]. Inflammation is typically neutrophilic and may be associated with fibrinonecrotic debris adhered to the intestinal lumen.
Pathogenic E. coli diagnosis is based on aerobic culture of either feces or intestinal tissues. At 37°C, E. coli grows readily on blood agar and on MacConkey agar, where it is lactose-fermenting. Because it is a normal ferret intestinal flora component, subsequent PCR-based testing should be used to confirm the presence of virulence determinants. Due to the broad range of potentially pathogenic ferret E. coli isolates that have been obtained, PCR tests for intimin, shiga-like toxins 1 and 2, heat-stable and heat-labile enterotoxins, and cytotoxic necrotizing factors 1 and 2 should be performed. Toxin liberation can be verified by adding liquid E. coli culture supernatant to cultured cells. Mannose-independent bacterial HEp2 and HeLa cell adherence has also been used as an adjunct E. coli virulence assay for human clinical isolates [84]. Because a variety of E. coli serogroups infect ferrets, serologic typing appears inferior to molecular diagnostics in determining E. coli isolate virulence. However, serotyping may be useful as an adjunct to other methods, such as pulsed-field gel electrophoresis and repetitive sequence-based PCR, in determining isolate relatedness during epizootics. E. coli infection must be distinguished from other causes of enterocolitis in ferrets, including Campylobacter, Salmonella, and L. intracellularis.
Because of the rapid-onset septicemia seen in many ferrets, empirical treatment should be instituted when colibacillosis is suspected. Antibiotics should be given immediately and adjusted based on culture sensitivity. Fluid losses and electrolyte abnormalities from diarrhea should be corrected through early isotonic crystalloid therapy.
A new species, Corynebacterium mustelae, was recently isolated from a septicemic ferret's lung, liver, and kidneys [85]. This species is closely related to Corynebacterium pseudotuberculosis and when cultured displays a distinct green-to-beige pigment and “humid cellar” odor. Corynebacterium ulcerans has also been isolated from research ferret cephalic implants and from 1 oropharyngeal sample [86]. Several others have isolated Corynebacterium spp. from ferret gingiva and conjunctiva [87,88]. While this organism is not currently associated with a definitive clinical disease entity, it should be considered in oropharyngeal disease, septicemia, implant infection, and mastitis cases. Toxin-producing Corynebacterium causes human diphtheria, an oronaspopharyngeal disease often characterized by a swollen neck and mucosal diphthoritic membrane formation. It has also been associated with bovine mastitis and subsequent transmission through unpasteurized milk consumption. While largely controlled in humans through vaccination, this remains a potential zoonotic hazard to ferret caretakers.
Historically, the normal stomach was considered bacteriologically inert because of the bactericidal effect of low gastric pH. During the past 25 years, studies in humans have shown that a newly recognized bacteria, Helicobacter pylori, causes gastritis, is causally associated with peptic ulcer disease, and is linked epidemiologically to gastric adenocarcinoma and mucosa-associated lymphoid tissue (MALT) lymphoma [89].
In 1985, a gastric Helicobacter-like organism was isolated from the margins of a ferret's duodenal ulcer [90] and was named H. mustelae [91]. Subsequently, in the United States, gastritis and peptic ulcers have been routinely reported in H. mustelae-colonized ferrets [92,93]. Virtually, every ferret with chronic gastritis is H. mustelae-infected, whereas uninfected specific pathogen-free (SPF) ferrets do not have gastritis, gastric ulcers, or detectable immunoglobulin G (IgG) antibodies [93–95]. Koch's postulates have been fulfilled by oral H. mustelae inoculation of naive ferrets. The resulting infection induced a chronic, persistent gastritis similar to natural infection [96]. H. mustelae has been isolated from ferret stomachs in England, Canada, and Australia but not from ferrets in New Zealand [97,98]. The mild gastritis noted in England may reflect sparse H. mustelae colonization in younger animals or reduced pathogenicity compared with American strains [98].
H. mustelae colonizes nearly 100% of ferrets shortly after weaning. Weanling and adult ferret feces have been screened to determine whether fecal transmission could explain this high prevalence [92,99]. H. mustelae was isolated from 8 of 74, 9-week-old ferrets, and 3 of 8, 8-month-old ferrets. These results were based on biochemical and phenotypic criteria as well as DNA probes and 16S rRNA sequencing [99]. H. mustelae was not recovered from the feces of 20-week-old ferrets that were positive at weaning or from the feces of l-year-old ferrets. However, given that H. mustelae persistently infects the ferret stomach, it is likely persistently shed. Further studies are needed to confirm this hypothesis.
During experimental ferret H. mustelae infection, transient hypochlorhydria was observed approximately 4 weeks after infection [96]. This period coincided with heavy H. mustelae colonization of the fundus. In ferrets experimentally infected for longer than 8 weeks, the organism (as in most naturally infected ferrets) colonized the antrum in greater numbers [96]. To test whether hypochlorhydria enhanced fecal H. mustelae recovery, oral omeprazole was administered to adult ferrets to induce gastric hypochlorhydria [53,96]. H. mustelae was isolated from sequential fecal samples in 23 of 55 ferrets (41.8%) during omeprazole therapy [53]. However, when the same group was not receiving omeprazole treatment, only 6 of 62 ferrets (9.3%) were H. mustelae-positive. All gastric biopsy samples were H. mustelae-positive, and in four of five ferrets their restriction enzyme patterns were identical to fecal H. mustelae strains. These findings favor the hypothesis that hypochlorhydria promotes fecal-oral spread [53,96].
In our laboratory, H. mustelae-infected ferrets are usually asymptomatic. Those with endoscopically visible gastric or duodenal ulcers can be recognized clinically by vomiting, melena, chronic weight loss, and lowered hematocrit. Acute gastric bleeding episodes are also occasionally noted [100] (Fig. 21.3 and Fig. 21.4). Clinical signs in ferrets with H. mustelae-associated gastric adenocarcinoma have included vomiting, anorexia, and weight loss—all signs easily confused with gastric foreign body [101].
In addition, like H. pylori-infected humans, H. mustelae-infected ferrets have hypergastrinemia, another feature that may be directly related to ulcerogenesis [102]. The plasma gastrin responses in H. mustelae-infected and noninfected ferrets are consistent with those in H. pylori-infected and H. pylori-eradicated humans, respectively (Table 21.2) [103–108]. Specifically, H. mustelae-infected ferrets have a significantly greater and more sustained increase in meal-stimulated plasma gastrin [102].
Table 21.2. Plasma Gastrin Concentrations in Ferrets of the Study (Mean ± SEM)
Gastric Helicobacter-induced hypergastrinemia may be related to peptic ulcer disease pathogenesis in ferrets and other animals. The combination of parietal cell trophic effects and increased gastric acid secretion may lead to upper gastrointestinal mucosal damage and ulcer formation. Abnormal gastrin hypersecretion may be a key element of Helicobacter-associated peptic ulcer disease pathogenesis. Current proposed mechanisms of human H. pylori-induced hypergastrinemia include G cell stimulation by amines or H. pylori-produced peptides, G cell stimulation by inflammatory cell cytokines, or selective somatostatin cell destruction or inhibition leading to the loss of somatostatin-mediated gastrin inhibition [108,109]. Thus, hypergastrinemia is attributable to either the bacterium itself or antral inflammation. H. pylori eradication leads to hypergastrinemia resolution. Because antibiotic therapy has been used successfully to eradicate H. mustelae from ferrets [110], studies in H. mustelae-infected and H. mustelae-eradicated ferrets can be undertaken to further investigate the mechanisms inducing hypergastrinemia. Future experiments are needed to determine the time interval for peak meal-stimulated gastrin secretion in ferrets, as are studies to determine whether H mustelae eradication results in decreased plasma gastrin concentration similar to human H. pylori eradication [103–108,111–114].
Gastric histopathologic changes closely coincide topographically with H. mustelae [94]. Superficial gastritis of the stomach body is associated with localization on the mucosal surface but not in the pits. In the distal antrum, inflammation occupies the full mucosal thickness, consistent with the diffuse antral gastritis described in humans. In this location, H. mustelae is seen at the surface, in the pits, and on the superficial portion of the glands. In addition, focal glandular atrophy (a precancerous lesion) and regeneration occur in the proximal antral and transitional mucosa. Also, deep H. mustelae colonization is focally observed in affected antral glands (Fig. 21.5). The H. mustelae-associated gastritis in the distal antrum and oxyntic mucosa also closely resembles human diffuse antral gastritis, which (as in the ferret) is usually accompanied by superficial gastritis of the stomach body. These clinicopathologic findings often underlie the duodenal ulcer syndrome [115,116]. Interestingly, the ferret is the only domesticated animal where naturally occurring Helicobacter-associated ulcer disease has been described (Fig. 21.6 and Fig. 21.7). The proximal antral and the transitional zone gastritis represent early-stage human multifocal atrophic gastritis, the entity underlying the gastric ulcer and gastric carcinoma syndromes [115,116]. The gastric disease severity increases as animals age (Table 21.3) [93].
Table 21.3. Severity of H. mustelae-Associated Gastritis Related to Ferret Age
Ultrastructural examination of ferret antral gastric tissue show that H. mustelae localizes within the gastric pits with little evidence of bacteria on the external surface or in the overlying mucus layer [116] (Fig. 21.8). H. mustelae are instead very closely associated with the epithelial cells. Organisms are present alongside microvilli perpendicular to epithelial cell surfaces and in some instances are seen penetrating these cells. Extensive microvilli loss and adhesion pedestals are visible. Occasionally, H. mustelae organisms are localized intraepithelially within membrane-bound inclusions, indicating their endocytic uptake. A dense fibrous-like matrix or glycocalyx exists between epithelial cells and H. mustelae when the two surfaces are very close together. H. mustelae is also present in intercellular junctions, but large bacterial numbers are not noted, unlike in human H. pylori infection [117,118]. Thus, close H. mustelae attachment may play a role in ferret peptic ulcer disease development as in H. pylori infection.
In vitro adhesion studies have shown that H. mustelae, like H. pylori, has specific ligand interaction with host lipid receptors [119,120]. Likewise, studies of H. mustelae and H. pylori cell surface properties indicate that they possess both hydrophobicity and hydrophilicity in an assay-dependent manner [121]. Studies imply, however, that the H. mustelae cell surface is relatively hydrophilic with distinct hydrophobic domains [121]. The ferret stomach is more hydrophobic than the large or small intestine. Interestingly, H. mustelae-infected inflamed gastric mucosa has a reduced mucosal hydrophobicity, which is consistent with human H. pylori infection [122]. Reduction of gastric epithelial mucus layer hydrophobicity may be responsible for this phenomenon [122]. These features may contribute to the pathogenesis of H. mustelae.
An isogenic urease-negative H. mustelae mutant was constructed to investigate urease's role in gastric mucosa colonization [123]. All four ferrets given the mutant strain remained uninfected throughout the study. Wild-type H. mustelae-infected ferrets exhibited diffuse mononuclear inflammation in the subglandular region and lamina propria of the gastric mucosa, whereas uninfected ferrets showed no or minimal inflammation. Thus, the inability of urease-negative H. mustelae to colonize the ferret gastric mucosa confirmed germ-free piglet studies in which isogenic urease-negative H. pylori mutants failed to colonize the stomach [124]. These studies proved urease's importance in gastric Helicobacter pathogenesis.
H. mustelae, like H. pylori, possesses two flagellin molecules, FlaA and FlaB [125,126]. To determine whether one or both proteins is necessary for pathogenesis, isogenic H. mustelae mutant strains were constructed where either or both genes were disrupted. Ferrets were given the FlaA (moderately motile), FlaB (weakly motile), or FlaA/B (nonmotile) mutant strain, the wild-type parent strain, or sterile broth [127]. Three-month infection with the H. mustelae wild-type parent strains and the FlaB mutant strain produced a mild lymphocytic and lymphofollicular gastritis, whereas infection with the FlaA mutant strain produced only minimal to mild mononuclear cell gastritis. The double mutant strain did not colonize, whereas the FlaA and FlaB single mutant strains initially colonized the ferret stomach at low levels, established a persistent infection, and generated increased H. mustelae numbers over time. Also, gastritis severity correlated with gastric bacterial numbers. Therefore, flagellar-induced motility, like urease, is an important H. mustelae virulence factor [127]. The publication of the H. mustele genome will further enhance study of its virulence properties [128].
Because H. pylori-associated gastric adenocarcinoma remains an important cause of human morbidity and mortality, it is important to consider the gastric cancer susceptibility of Helicobacter-infected animals [53,129–131]. We documented argyrophilic bacteria, consistent with H. mustelae in location and morphology, within the pyloric mucosa of two male ferrets with pyloric adenocarcinoma [101]. In another pyloric adenocarcinoma case, argyrophilic bacteria with H. mustelae-like morphology were also seen within pyloric pits, and the neoplasms manifested as multiple foci of tubules lined by mucous epithelium that invaded into the deep submucosa, resulting in a significant local scirrhous response [132]. The discrete invasive characteristics suggested that the neoplasms represented early cellular infiltration beyond carcinoma in situ. The neoplastic growth patterns resembled the description of three naturally occurring pyloric adenocarcinomas previously reported [133,134]. This pattern was also seen in N-methyl-N-nitro-N′-nitrosoguanidine-induced gastric adenocarcinomas in H. mustelae-infected ferrets [53]. The young adult ages in this (2 and 3.5 years) and other (3 to 4 years) cases suggest early disease progression (Fig. 21.9) [133,134].
More confluent, highly anaplastic, and transmurally invasive adenocarcinomas, consistent with protracted tumor progression, were also reported in an older ferret [132]. Osseous metaplasia was noted in the tumors, providing a useful marker during radiographic examinations [53,133,134] (Fig. 21.10). Unlike the previous reports that cited poor prognosis associated with ferret gastric adenocarcinoma, both ferrets in the recent report responded well to surgical intervention and remained asymptomatic postoperatively for 1 year and 6 months, respectively [132]. Thus, the presence of naturally occurring H. mustelae-associated gastric adenocarcinoma supports the epidemiologic association between H. pylori infection and gastric cancer risk in humans [135–139]. Additionally, H. pylori-associated MALT lymphoma has been recently described in humans, and we have recently documented several H. mustelae-infected ferrets with the same syndrome [140] (see Chapter 9).
We also have noted possible Helicobacter-associated cholangiohepatitis and neoplasia in a group of co-housed pet ferrets. Eight of 34 (24%) of ferrets, all 5- to 8-years old, exhibited chronic cholangiohepatitis and bile duct hyperplasia, as well as variable oval cell hyperplasia, cystadenoma, and hepatocellular or cholangiocellular degenerative change. Cholangiocellular carcinoma was identified in two of eight ferrets. Spiral-shaped argyrophilic organisms, sharing sequence similarity with H. cholecystus, were identified in three ferrets, including both animals with carcinoma. While co-housing supports an infectious etiology, the precise role of Helicobacter spp. in ferret hepatobilliary disease remains poorly defined [141] (see Chapter 16, “Gastrointestinal Diseases”).
Ferret H. mustelae isolates have biochemical features similar but not identical to H. pylori, particularly in regard to abundant urease production (Table 21.1). For culture, gastric biopsy or necropsy tissues are transported in sterile phosphate-buffered saline and processed within 1 hour of collection. Gastric samples are minced with sterile scalpel blades and inoculated onto blood agar plates supplemented with trimethoprim, vancomycin, and polymixin B (Remel, Lenexa, KS). The plates are incubated at 37°C or 42°C in a microaerobic atmosphere (80% N2 10% H2 and 10% CO2) for 3–7 days. Bacteria are identified as H. mustelae on the basis of Gram stain morphology, urease, catalase, and oxidase production, cephalothin resistance, and nalidixic acid sensitivity (Table 21.1).
A provisional gastric Helicobacter diagnosis takes advantage of their unique ability to produce large quantities of urease. Gastric biopsy specimens containing bacteria can be placed in urea broth containing a pH indicator (phenol red) and a preservative (sodium azide) to help prevent false-positive reactions due to urease-positive bacterial contaminants. The broth assay is then visually monitored over 16 hours. A deep pink color change, caused by the enzymatic breakdown of urea to ammonia, indicates a positive reaction. Also, a semiquantitative assessment of H. mustelae numbers can be inferred by color change speed. A test is available commercially, but a microtiter tray with a measured amount of urea test solution delivered to each well can be effectively and economically used. The urea breath test, which measures expired radiolabeled carbon dioxide consumed in a test meal, is used in humans but has only been perfected under experimental conditions in ferrets [142].
Although infected animals and humans mount a significant systemic IgG response to gastric Helicobacter, the immunoglobulins are not protective. Nonetheless, a serologic assay is being used to diagnose human H. pylori and ferret H. mustelae infections. This enzyme-linked immunosorbent assay provides a reliable noninvasive diagnostic test for H. mustelae infection [94]. An H. pylori detection assay performed on stool is routinely used to diagnose human disease, and a comparable ferret test might prove useful in H. mustelae diagnosis.
Triple therapy consisting of oral amoxicillin (30 mg/kg), metronidazole (20 mg/kg), and bismuth subsalicylate (17.5 mg/kg, Pepto-Bismol original formula, Proctor & Gamble, Cincinnati, OH) 3 times a day for 3–4 weeks has successfully eradicated H. mustelae from ferrets [110]. In addition, a treatment regimen used to eradicate human H. pylori has been successfully used to eradicate ferret H. mustelae [143]. Ferrets received 24-mg/kg ranitidine bismuth and 25-mg/kg clarithromycin orally 3 times daily for 2 weeks. H. mustelae cultured gastric biopsy specimens obtained at 16, 32, and 43 weeks after treatment termination indicated that long-term eradication was achieved in six of six ferrets. With eradication, gastritis diminished and H. mustelae-specific IgG antibody decreased. These results are consistent with findings in humans following H. pylori eradication.
In ferrets, daily oral omeprazole at 0.7 mg/kg effectively induces hypochlorhydria [99] and may be used in conjunction with antibiotics to treat H. mustelae-associated duodenal or gastric ulcers. Oral cimetidine (10 mg/kg 3 times a day) can also be used to suppress acid secretion. Acute bleeding ulcers must be treated as emergencies, and fluid and blood transfusions are essential.
Leptospira is a Gram-negative spirochete transmitted via urine-contaminated water sources. Leptospirosis in animals was first diagnosed in 1850, 30 years before recognition of the human disease. Several Leptospira serovars have been isolated from ferrets [144], and leptospires have been postulated as a zoonotic threat [145].
The historic use of ferrets in domestic and farm rodent control likely exposed them to leptospires shed in rodent urine. Fur-bearing animals maintained in commercial operations, including ferrets, can contract leptospirosis. The incidence in captive fur bearers is relatively low, but disease may occur either as an enzootic with isolated individual cases or as an epizootic [144,146]. Given the high disease incidence in domestic animals, particularly dogs, pet ferrets are probably at risk, although ferret leptospirosis reports are fragmentary.
Wild mustelids, including nine stoats (Mustela erminea), nine ferrets (Mustela putorius), and four weasels (Mustela nivalis) inhabiting farmland on New Zealand's north island were surveyed for leptospiral infection. None had evidence of infection by serology (using 12 different leptospiral antigens) or kidney culture [147]. Despite a high prevalence of enzootic L. ballum infection in house mice (Mus musculus), ship rats (Rattus rattus), and other New Zealand rat populations and their status as a major Mustelid sp. food source, infection with this serotype was not demonstrated. In Denmark, 10% of stoats examined had serologic titers to serovars Leptospira pomona and Leptospira sejroe. However, 11 polecats (M. putorius) and 16 weasels examined were negative [148]. In Great Britain, a similar survey indicated that weasels have had serologic titers to serovars L. sejroe and Leptospira bratislava [149,150]. Another survey examined farm-raised and wild mink, as well as European polecats, in southwestern France [151]. Slide agglutination testing indicated a high Leptospira antibody prevalence that was host species-dependent. Antibody prevalence was 65–86% in wild European mink, European polecats, and American mink. Leptospira serovars Australis and Icterohemorragiae were most common across wildlife species, though Leptospira grippotyphosa, L. sejore, and L. panama were also frequently noted. In farmed American mink, seroprevalence was 31%, and included all of these five serovars except L. panama. Interestingly, 18% of farmed American mink were seropositive for Leptospira autumnalis, which was not frequently noted in wildlife. PCR testing found that 15–23% of wild mustelid spp. were positive for renal carriage. That renal prevalence is only one-third of seroprevalence suggests that urine shedding may be transient, though such renal prevalences, if they correlate well with shedding, could indicate high infection risk.
Precise health risks to affected ferrets remain enigmatic, and likely would consist of spontaneous abortions and renal damage that would threaten individual health and longevity. However, Leptospira spp. remains an important concern because of its zoonotic potential and ability to infect other domestic animals, particularly pet dogs. Important control mechanisms include rodent control and limiting access to wildlife and potentially contaminated water sources. A properly produced vaccine, if warranted, should protect ferrets against clinical disease and development of a carrier state. To date, however, leptospirosis vaccines have not been recommended for commercial, pet, or research ferrets.
Listeriosis is a mildly infectious but highly fatal disease that affects several animal species, including humans [152]. The etiologic agent, L. monocytogenes, is a gram-positive, non-spore-forming rod that is motile at room temperature and hemolytic, characteristics that separate it from other similar diphtheroid bacteria. It is usually isolated from the CNS of symptomatic animals or from infected tissues. It is occasionally isolated from asymptomatic hosts.
The organism's prevalence in ferrets is unknown. It was isolated from ferrets previously inoculated with a suspension of bacteriologically sterile lung tissue from distemper-infected ferrets [150]. Their lungs and spleens were subsequently found to harbor L. monocytogenes, which was identified both by culture and by intracerebral hamster inoculation [153]. L. monocytogenes has also been isolated from another mustelid, the sable (Mustela zibellina) [154]. It has also been isolated from the pleural fluid of an immunosuppressed ferret with adrenal cortical disease and cardiomyopathy [14]. Pneumonia and hepatitis in this animal were attributed to L. monocytogenes infection.
Contaminated food ingestion and aerosol inhalation are its suspected modes of transmission. The organism may be isolated from water, soil, dust, animal feed, and various domestic and wild animals. It is not known whether ferret in utero infection occurs as it does in humans. Asymptomatic ferrets may play a role in disease dissemination, but experimental evidence concerning this hypothesis is lacking. However, asymptomatic L. monocytogenes fecal carriers have been reported in both humans and animals.
L. monocytogenes-associated ferret clinical signs and pathology have not been described, and are presumed to be the same as those encountered in other species. Accordingly, they are likely to be CNS-confined or septicemia-associated. The disease is frequently superimposed on other debilitating diseases, especially in patients receiving steroid therapy or other immunosuppressive drugs [4,14].
Ferrets presenting with CNS signs should be quarantined and considered rabies-suspect. Therefore, the animal should be handled minimally and with extreme caution. If rabies exposure can be definitely excluded, a cerebrospinal fluid culture should be performed. Listeria grows aerobically and tests positive for catalase and the Voges–Proskauer reaction. It is important not to misidentify it as a diphtheroid culture contaminant. Cold enrichment techniques may be used for increased culture sensitivity.
Treatment consists of the use of antimicrobial agents, penicillin and ampicillin. In vitro, L. monocytogenes isolates are also usually sensitive to chloramphenicol, tetracycline, erythromycin, and cephalothin.
The ferret appears to be highly susceptible to certain Mycobacterium species of avian, bovine, and human origin. Disease may be subclinical or produce nonspecific clinical signs and therefore escape detection, thus potentiating disease spread among animals and zoonotic transmission to pet owners or laboratory personnel.
Only scattered reports of ferret mycobacterial infection appear in the literature, almost exclusively in English and European research ferrets from 1929 to 1953 [155,156]. One investigator reported that among thousands of necropsies performed on dealer-obtained ferrets, 60% were infected with Mycobacterium—most commonly avian strains and, less frequently, bovine strains [157,158]. However, a wildlife survey utilizing pooled lymph nodes from more than 22,000 mustelid spp. identified 476 Mycobacterium isolates [159], 56% and 44% of which were Mycobacterium bovis and Mycobacterium avium complex, respectively. Forty-eight M. bovis isolates were subjected to restriction enzyme analysis, which identified 23 unique types, most of which are associated with both domestic animals and wildlife. The disease continues to be recognized in New Zealand ferret farms [160]. Atypical mycobacterium infection due to M. avium has been recently diagnosed in an American pet ferret [161].
The ferret can be naturally or experimentally infected with bovine, avian, and human Mycobacterium species. Historically, ferret infection was recorded when feeding unpasteurized milk, raw poultry, and meat (including meat by-products) was routinely performed. Feeding of infected carcasses or milk likely contributed to disease incidence during this era. Chicken offal used in preparing mink ration, which is also fed to ferrets, can be a possible infection source [162]. Efforts to eliminate tuberculosis in commercially reared livestock and chickens have effectively reduced this disease's incidence. In addition, the general use of commercially prepared cat or ferret diets as sole ferret food sources has appreciably reduced the possibility of disease introduction. Mathematical models suggest that transmission within wild ferret populations is also primarily through food contamination [163]. This was hypothesized because age-prevalence curves indicated a sharp increase in infection following weaning that remained throughout life. Higher male disease prevalence was accounted for by factors such as home range size, susceptibility, and dietary composition. Additionally, wild birds, which can shed avian mycobacteria in their feces, may contaminate feed supplies and outdoor ferret housing areas. M. bovis has also been transmitted from experimentally infected ferrets to 75% of co-housed, noninfected ferrets [164]. It was suggested that factors such as playing, fighting, cannibalism, mating, and fecal exposure may have been involved.
Clinical signs exhibited by infected ferrets are poorly documented, highly variable, and nonspecific. Systemic infection caused by bovine strains results in generalized weight loss, appetite loss, lethargy, and death. In one well-documented tuberculosis case caused by a bovine-origin strain, the ferret had hind limb paralysis that began with difficulty walking and progressed to hind limb splaying. Bacteria isolated from this animal were inoculated intramuscularly into another ferret, which remained normal until 6 months later, when it developed paralysis of the adductor thigh muscles. Eventually, all muscles of the limbs were affected [157]. Splenomegaly, hepatomegaly, and intestinal nodules can also be palpated in disseminated disease.
Disseminated disease is more likely with M. bovis, whereas the human and avian tubercle strains usually produce only local, indolent tubercular lesions. In one report, nine experimentally M. bovis-infected ferrets had macrophage infiltration most frequently in the liver [164]. Infiltration was also variably noted in the mesenteric, retropharyngeal, bronchial, and mediastinal lymph nodes. Acid-fast bacilli were noted in the thoracic lymph nodes in five of nine ferrets, but were not identified in the liver or mesenteric lymph nodes.
In vitro phytohemagglutinin (PHA) responses of normal versus M. bovis-treated ferret peripheral leukocytes demonstrate that M. bovis PHA suppression correlates with M. bovis peripheral leukocyte cytotoxicity. M. bovis concentrations 106/mL and below enhance the PHA response, but more than 106 organisms/mL suppresses it [165]. Thorns and Morris [165] speculated that the cytotoxic activity of high M. bovis concentrations and associated in vitro PHA-stimulated leukocyte suppression may play a role in disease pathogenesis by depressing specific and nonspecific cell-mediated immunity. This explains the lack of cellular tissue reactions seen in naturally occurring M. bovis ferret disease.
In another case, a pet ferret presented with intermittent anorexia, diarrhea, and vomiting of 6 weeks' duration [161]. Supportive care was given for another 3 weeks without resolution of signs. Radiography with barium indicated delayed gastric emptying. A midline exploratory celiotomy was performed, a 1 cm constricted band was located in the proximal jejunum, and the intestine was dilated proximal to the constriction. A resection and anastomosis was performed, and the resected tissue revealed granulomatous jejunal inflammation. Diffuse macrophage infiltration with intracellular acid-fast organisms was also observed in a mesenteric lymph node biopsied during surgery [161]. The ferret responded to the surgery and did well for the next 8 months, when vomiting and weight loss recurred. Abdominal surgery revealed a 1-cm diameter pyloric mass that projected into the gastric lumen. This mass was resected, but despite surgical intervention and treatment, the animal's condition deteriorated and it died 2 weeks later. Necropsy revealed a thickened pylorus. Multiple tissues were submitted for microscopic analysis, and liver and spleen sections were cultured. Histologically, sheets of macrophages were seen in the lamina propria and submucosa. Severe granulomatous inflammation was also noted in the jejunum, stomach, and mesenteric lymph nodes. Additionally, small granulomas were noted in the liver and spleen. Intracellular acid-fast organisms were present in the jejunum and elsewhere, and M. avium was subsequently cultured from the liver and spleen.
M. avium was also reported in a 6-year-old pet ferret with chronic weight loss accompanied by leukocytosis and anemia [166]. Radiography revealed hepatomegaly and a left cranial abdominal mass. Because metastatic neoplasia was suspected, euthanasia was elected. Upon necropsy, mesenteric lymph node enlargement, multifocal tan hepatic nodules, and a thickened stomach were evident. Both neoplastic cells and granulomatous inflammation were evident in these locations, and acid-fast bacilli were present within macrophages. Non-paratuberculosis M. avium was identified by PCR, and it was hypothesized that large cell lymphoma-mediated immunosuppression facilitated disseminated mycobacterial infection.
Granulomatous enteritis with acid-fast organisms present in hepatic, colonic, and mesenteric lymph node lesions, as well as fecal smears, has also been reported in two laboratory ferret groups with diarrhea of several months' duration. Unfortunately, the bacteria were not cultured or identified [167]. Pulmonary M. avium has also been reported in ferrets maintained in a French zoological garden [168].
Mycobacterium celatum, which infects immunosuppressed humans, has been identified in three ferrets [169–171]. In these cases, 3- to 5-year-old ferrets presented with 2- to 6-month histories of weight loss with variable coughing, depression, muscle atrophy, vomiting, and mild diarrhea. Multifocal pulmonary nodules were evident radiographically, and euthanasia was elected in all cases. On necropsy, firm 2- to 10-mm, pale gray to light brown nodules were randomly distributed throughout the lung parenchyma. Lymphadenopathy and splenomegaly were variably evident, as were smaller 1- to 2-mm pale hepatic foci. In two cases, granulomatous lesions were evident microscopically, and acid-fast bacteria were located within lesions and inside macrophages. In the third case, antemortem ultrasound-guided fine needle aspirates identified the bacterium within macrophages. While no mycobacteria were cytologically evident, chronic systemic antimicrobial therapy was unsuccessful, and its discontinuation precipitated decline and death. In all three cases, culture followed by 16S rRNA sequencing confirmed the etiologic agent.
Mycobacterium abscessus has been associated with pneumonia in two co-housed ferrets [172]. Again, weight loss and coughing were the primary symptoms; bronchoalveolar lavage (BAL) identified a primarily neutrophilic inflammatory response, and acid-fast bacilli were noted within macrophages. BAL fluid culture yielded bacterial colonies within 5 days, consistent with this organism's rapid growth. Another report describes Mycobacterium genavense infection associated with disseminated lymphatic infection and conjunctivitis in a ferret [173]. This is another saphrophytic Mycobacterium species associated with human AIDS patients. Lymph node cytology and conjunctival biopsy, followed by staining and 16S rRNA sequencing, confirmed infection.
Nodular lesions, if calcified, can be demonstrated radiographically. Laparotomy, with biopsy of involved abdominal lymph nodes and other organs and bacterial isolation, will establish a definitive diagnosis. M. avium was isolated from ferret tissue by macerating infected tissue in a cetylpyridinium chloride solution (7 g/L of distilled H2O, Sigma Chemical, St. Louis, MO) for 3 hours. Following centrifugation at 300× g for 15 minutes, sediment was inoculated onto Herrold's egg yolk media slants and incubated for 3 weeks at 35° and 42°C [161]. Additionally, PCR-based M. tuberculosis diagnosis is being used with increasing frequency in humans and has also been utilized to confirm and speciate ferret infections [166]. Immunohistochemical reagents are also available, though species specificity may not be reliable [169].
Experimentally, ferrets inoculated with Freund's complete adjuvant (FCA, containing killed M. tuberculosis) react to an intradermal 104 U (200 μg) tuberculin injection (see Chapter 7) [174]. In other studies, ferrets inoculated with both FCA and 104 and 105 U of tuberculin had skin reactions 14 days later [175]. Skin tests carried out 4 weeks after this treatment were negative. However, in ferrets experimentally infected with M. bovis, a minimal tuberculin response to purified protein derivative (10 μg/mL) was observed only 36 hours after inoculation, and no response was observed at 7 and 15 days after inoculation [176]. The tuberculin skin test was used to eliminate natural tuberculosis cases in a ferret breeding colony [177]. Unfortunately, the tuberculin type and dosage were not detailed. Depending on the infecting Mycobacterium strain, reaction to a purified protein derivative tuberculin test may be minimal. For example, Pulling [178] found that mink from a tuberculous colony tested by intradermal bovine tuberculin injection did not show a delayed hypersensitivity reaction.
Because of the zoonotic risk associated with tuberculosis, M. tuberculosis- or M. bovis-infected ferrets should be euthanized. Although M. avium is not a disease reportable to public health officials, its presence may present a zoonotic risk, particularly in immunocompromised humans. If other ferrets have been exposed, they should be tuberculin-tested (which may not be helpful), or other diagnostic tests should be performed to ascertain their disease status. Personnel at risk should also be screened with appropriate tuberculin tests.
Recurrent M. bovis infection following a ferret bite has been recorded [179]. In 1991, a 63-year-old man presented with a 3-week history of right palm swelling that became red, tender, and painful. Pyrazinamide-resistant M. bovis was isolated. Subsequent treatment with ethambutol for 2 months, together with rifampin and isoniazid for 7 months, led to an uneventful recovery. The patient had been bitten by a ferret 22 years earlier. The chronic swelling situated above the carpus badly impaired finger mobility. Histologically, the patient had tuberculous synovitis, and M. bovis was isolated. Previous therapy with streptomycin sulfate with supplemental para-aminosalicylic acid and isoniazid, although apparently clinically successful, had not completely eradicated the bacterium, and subsequent infection reactivation occurred [179]. Molecular evidence also supports M. bovis human–ferret transmission [180]. In one study, 43% and 26% of human- and ferret-origin isolates, respectively, shared restriction enzyme analysis (REA) patterns with isolates from the other species. However, population modeling suggests that wild ferrets residing in New Zealand do not exist at sufficiently high density to act as sole M. bovis maintenance hosts, emphasizing the importance of other wildlife species in this dynamic [181].
Mycoplasma spp. were first isolated from clinically normal ferret oral and nasal cavities in Japan [182]. Another organism, Mycoplasma mustelae, has been isolated from healthy kit lungs on three Danish mink farms. Its pathogenic potential awaits further study [183]. Recently, mycoplasmosis has been recognized as an emerging ferret respiratory disease. In the initial report, approximately 8000, 6–8 week old ferrets presented with conjunctivitis, a dry (honking) cough, and dyspnea [184]. White nodules were grossly apparent on the lung, and peribronchial lymphocytic infiltration with bronchial-associated lymphoid tissue (BALT) hyperplasia was microscopically identified. PCR and culture of BAL samples were used to achieve a diagnosis. It is noteworthy that nasal and conjunctival swabs were nondiagnostic. The precise species involved was uncharacterized and shared greatest similarity with Mycoplasma molare and Mycoplasma lagogenitalium. Mortality was high only in kits, and the infection's source was not determined.
Because ferret mycoplasmosis produces clinical signs and lesions similar to mycoplasmosis in other species, similar treatment should be considered. Tetracycline is typically used to treat infected pigs and rats [185,186]. However, antibiotic treatment fails to completely eliminate bacteria in these species, necessitating vaccination or husbandry changes to prevent reoccurrence.
Proliferative bowel disease was first reported in research ferrets in 1983 [41] and in a pet ferret in 1986 [187]. It was initially classified as a Campylobacter-like organism similar to those noted in hamsters, rabbits, and pigs. The organism was formally named L. intracellularis in 1995 and is now recognized as a common clinical entity in weanling ferrets.
Proliferative bowel disease was originally described in a research ferret colony, where 31 of 156 (20%) ferrets developed protracted diarrhea over a 4-month period [41]. Ten of these were examined by necropsy and histology and were found to have proliferative bowel disease. Although C. jejuni was isolated from 6 of 10 ferrets, the causal organism was an intracellular Campylobacter-like organism closely related to Desulfovibrio spp. [41]. L. intracellularis, as it is now classified, is appreciated as an important cause of intestinal disease in many species, including ferrets, pigs, hamsters, horses, rabbits, primates, and deer. Isolates from these species appear closely related, and the ability to reproduce the pig disease syndrome in hamsters and mice suggests a low degree of host restriction [188,189]. Exposure to other species known to harbor L. intracellularis should be considered a potential source of infection. It is also possible that interaction with other gastrointestinal bacteria modulates disease, as L. intracellularis-inoculated, restricted flora piglets fail to become colonized or develop disease [190]. Similar piglets receiving L. intracellularis and other enteric bacteria reproduce typical proliferative enteropathy [191]. Additionally, simultaneous infection with L. intracellularis and enteropathogenic E. coli (EPEC) has been characterized in rabbits, and synergism associated with high mortality was suggested [192]. Little is known regarding precise transmission mechanisms, though fecal-oral spread is suspected. It is thought that environmental sanitation and reduced exposure to adult ferret feces, coupled with maintenance of a high nutritional plane in young ferrets, may reduce disease severity.
In the original disease description, 10 animals ranging in age from 4 to 6 months presented with diarrhea, weight loss, and rectal prolapse (Table 21.4) (Fig. 21.11. Fecal material was usually blood-tinged, often discolored green, and contained mucous. Intermittent diarrhea persisted for more than 6 weeks. Seven of the 10 animals exhibited more than a 100-g weight loss during the first 2–4 weeks of illness. Affected males lost 12–54% of their body weight, and affected females lost 30–54%. Several ferrets exhibited ataxia and muscular tremors. Similar clinical presentations were noted in pet ferret [187].
Table 21.4. Ferrets with Proliferative Colitisa
Clinical data | Number of affected ferrets |
---|---|
Age at onset of diarrhea 4–6 months | 10 |
Partial rectal prolapse | 10 |
Diarrhea | |
Mucohemorrhagic | 7 |
Greenish, mucoid | 3 |
CNS signs: ataxia, muscular tremors | 5 |
Weight loss, >100 g in 2–4 weeks | 8 |
Rectal cultures | |
Campylobacter jejuni | 6 |
Salmonella species | 0 |
Sex incidence | |
Males | 7 |
Females | 3 |
Disposition of animals | |
Died | 4 |
Euthanized | 6 |
Duration of illness | |
6–18 weeks | 7 |
Less than 3 weeks | 3 |
aTen ferrets seen clinically were used in this study.
Gross morphologic changes consist of a palpable segmented thickened lower bowel, usually the terminal colon, with associated histologic changes (Fig. 21.12). The ferret does not have a cecum, and thus lacks a readily identifiable division between the small and large intestines. Limited studies indicate that the transition from villous to nonvillous mucosa occurs in the area analogous to the human transverse colon. Histologically, ferret proliferative bowel disease is defined by marked mucosal cell proliferation and intracytoplasmic L. intracellularis clustered on the apical epithelium. The lesions observed in the colon (and less frequently in the ileum) mimic the proliferative changes described in the hamster ileum, often referred to as wet tail, proliferative ileitis, atypical ileal hyperplasia, hamster enteritis, and enzootic intestinal adenocarcinoma [193–196]. The histology also mimics the proliferative lesions described in the swine ileum, again referred to as various syndromes, including intestinal adenomatosis, adenomatous intestinal hyperplasia, proliferative ileitis, and proliferative hemorrhagic enteropathy [197,198]. Lesion location is the primary difference in ferrets when compared with hamsters and pigs. In ferrets, lesions are primarily found in the colon, while in the hamster and pig, the lesions are usually in the ileum with occasional extension to the colon and jejunum [194,196]. In ferrets, the microscopic epithelial hyperplasia, glandular irregularity, reduced goblet cell production, and variable inflammatory cell infiltrate (in type and severity) are consistent with hamsters and pigs (Fig. 21.13). L. intracellularis and intestinal gland herniation through the muscularis mucosa into the submucosa and tunica muscularis is commonly reported in hamsters [193,194]. This feature was observed to some degree in all ferrets. Intestinal gland translocation into regional lymph nodes and liver can also occur [199] (Fig. 21.14 and Fig. 21.15). Gross and microscopic findings indicate an increased tunica muscularis thickness possibly due to muscle fiber hypertrophy, although this has not been confirmed.
Grossly, an enlarged, palpable colon is evident. Biopsy of affected tissue reveals hyperplastic mucosa, and silver stain demonstrates the organism within the apical epithelium [200] (Fig. 21.16). In selected cases, calcification of involved extraintestinal tissue is noted radiographically [201]. Fluorescent antibody techniques have demonstrated L. intracellularis antigen within hyperplastic ferret colonic epithelial cells [200] (Fig. 21.17). The same antigen is present within the affected epithelia from porcine intestinal adenomatosis and hamster proliferative ileitis cases [201].
The immunoperoxidase monolayer assay (IMPA, a serologic test) and fecal PCR are the most commonly used L. intracellularis diagnostic tests. In pigs, IMPA has been shown to be highly sensitive and specific [202]. Several fecal PCR assays have been reported, all with high specificity but variable sensitivity, most likely due to cyclic bacterial shedding [203,204]. These PCR assays could be adapted for postmortem evaluation during epizootics. An accurate antemortem diagnosis is perhaps best accomplished using both IMPA and fecal PCR, as has been proposed in suspect horses [205]. These results should be interpreted along with clinical signs.
In the original disease description, two affected animals died after they were given broad-spectrum antibiotics (chloramphenicol or gentamicin) and fluid therapy. Two others died following 5 days of marked weight loss and diarrhea [41]. While six ferrets treated with supportive parenteral fluids and antibiotics during episodic diarrhea responded temporarily, they subsequently relapsed and were euthanized. More aggressive therapy with 25–50 mg/lb chloramphenicol intramuscularly, subcutaneously, or orally twice daily for 10–14 days resulted in clinical sign remission and histologic lesions regression [206]. Because oral chloramphenicol is not routinely available, compounded palatable chloramphenicol suspensions can be utilized instead. Others have had success using oral metronidazole at 20 mg/kg twice daily for 10–14 days [207]. Supportive therapy with subcutaneous fluids and oral caloric and amino acid supplementation (e.g., Nutrical, Evsco, Buena, NJ or Liquical, Butler, KY) is essential.
Although salmonellosis is seldom reported in the ferret literature, it most likely occurs with some frequency. It should be included as a differential in cases of gastroenteritis or abortion in which a microbial etiology is suspected. The fur industry has historically referred to the disease in ferrets and minks as food poisoning.
The genus Salmonella is composed of motile bacteria that confirm to the definition of the family Enterobacteriaceae. The nomenclature employed to describe the genus Salmonella has been confusing, given the use of multiple schemes in the literature and the historical practice of considering different serotypes of Salmonella to be different species [208]. The genus Salmonella is composed of two species, Salmonella enterica and Salmonella bongori. Salmonella enterica has been subdivided into six subspecies: S. enterica subsp. enterica, designated subspecies I. Subspecies I strains are commonly isolated from humans and warm-blooded animals. Subspecies II, IIIa, IIIb, IV, and VI strains, and S. bongori are usually isolated from cold-blooded animals and the environment [208].
The prevalence of salmonellosis in ferrets is unknown. Before its isolation from an asymptomatic animal's liver and spleen in 1947, the organism, according to the literature, had not been isolated from ferrets [209]. Its occurrence depends largely on exposure to different Salmonella serotypes present in raw meats and meat by-products. In a 9-month research ferret survey, 4 of 99 (4%) had Salmonella spp. isolated by fecal culture; the serotypes identified were Hadar, Enteritidis, Kentucky, and Typhimurium [210,211]. The practice of feeding uncooked meat was presumed to have caused the infection. This assertion is supported by a Salmonella Dublin epizootic in farmed mink and foxes [212]. In this case, 25 farms experienced an increased incidence of barren females, abortion, and stillbirth, as well as increased dam and kit mortality. All affected farms utilized the same feed provider, and contaminated raw bovine offal was thus causally implicated. This report also supports the assertion that mink kits and dams are exquisitely susceptible to salmonellosis during pregnancy and immediately post parturition.
Improperly processed animal pet foods may also be contaminated with Salmonella [213]. Morbidity and mortality rates vary, but one report cited a 10% morbidity rate. In this epizootic, Salmonella Typhimurium was isolated from two ferrets, one that died of enteritis and another with bloody diarrhea. The infection source was not identified. Because of its highly contagious nature, disease spread should be precluded by strict disinfection and handler personal hygiene. Clinically ill or asymptomatic ferrets may serve as human infection sources. Additionally, biosecurity measures, particularly cooked feeds, should be used to prevent pathogen introduction.
Infected ferrets can be asymptomatic Salmonella carriers or, depending on host resistance and infecting strain, may be clinically affected. Infected animals typically present with lethargy and gastrointestinal symptoms of varying severity. When experimentally fed to three ferrets on a marginal nutritional intake, Salmonella Newport produced lassitude, anorexia, and muscular trembling in one ferret the day after inoculation. This ferret had bloody feces on day 2 and died on day 3 [214]. Accompanying signs often seen with Salmonella-associated gastroenteritis include dehydration, anorexia, moderately elevated temperature, pale mucous membranes, and malaise. Although not documented in ferrets, cats with clinical salmonellosis can have a lowered white blood cell count, with significant neutropenia and lymphopenia [215]. Acute septicemias with thrombocytopenia and nonregenerative anemia are also occasionally noted.
In a similar experiment, two ferrets were inoculated with S. Typhimurium isolated from an asymptomatic ferret's spleen and liver [209]. The animals became febrile and emaciated and exhibited conjunctivitis with a clear watery discharge. One also had balanitis. Ten days after inoculation, one animal was moribund and both were euthanized. Both had tarry intestinal contents, and petechial hemorrhage was noted on one ferret's gastric mucosa. Although the authors stipulated that the temperature curve noted was not characteristic of distemper, they could not exclude the possibility of concurrent infection with other organisms [209]. Ferrets naturally infected with salmonellosis (uncomplicated by concurrent infection) also demonstrated severe conjunctivitis, tarry stools, and typhoid-like temperature fluctuations [216]. In another salmonellosis epizootic in biomedical research ferrets, 2 of 20 animals with blood-tinged diarrhea had S. Typhimurium isolated from their feces. In addition, cats and guinea pigs can similarly present with Salmonella-associated conjunctivitis [217,218].
Gross and microscopic findings are similar to those seen in other Salmonella-infected animals. Gross examination of experimentally or naturally infected ferrets reveals hyperemic intestinal serosal vessels, distended gallbladder, and small intestine filled with dark red, semisolid material. Histologically, the gastrointestinal tract has gastric mucosal congestion with a mucous layer consisting of inflammatory cells and desquamated epithelium [214]. The small intestine has a similar mucous layer exudate with tips of villi sloughed in the intestinal lumen. The small intestine may have marked macrophage and lymphocyte infiltration in the submucosa, lamina propria, and mucous epithelium.
Elsewhere, the liver, spleen, and mesenteric lymph nodes routinely contain necrotic areas, which can be detected grossly as yellowish-white foci of varying size. Although not pathognomonic, necrotic foci referred to as “paratyphoid nodules” are often visible on the liver's surface or in cut sections. Splenomegaly is also encountered. Thrombosis of abdominal vessels is sometimes encountered and may be consistent with disseminated intravascular coagulation. In cases of septicemia, ecchymoses and petechiae can be noted on the visceral and parietal pleura, peritoneum, endocardium, epicardium, and meninges.
Abortion due to Salmonella Choleraesuis has also been noted in mink fed infected raw pork livers [214]. S. Typhimurium-associated abortion has also been seen in mink fed raw beef offal [212]. In this case, clinical signs were primarily inappetence, abortion, stillbirths, and sudden death of both dams and kits. When 52 mink were necropsied, hyperemic and friable uteri were evident, often with endometrial rupture and purulent peritonitis. Histopathologically, this manifested as deep, necrotizing endometritis and myometritis. Uterine contents were thick, crimson, and odorless. While the cervix was typically not dilated, vaginal discharge was frequently evident. Less frequently, splenomegaly and pinpoint pulmonary hemorrhages were noted.
Salmonella is a gram-negative rod and grows readily in enrichment media, such as GN broth, selenite broth, and tetrathionate broth. Plating media used are Salmonella-Shigella agar, brilliant green agar, eosin-methylene blue agar, and MacConkey agar. The organism can be recovered from liver, heart blood, spleen, intestinal contents, and occasionally from bone marrow. Salmonella must be differentiated from other causes of ferret gastrointestinal disease, particularly Campylobacter spp., E. coli, and L. intracellularis.
It is important not only to recognize the clinical syndrome of salmonellosis but also to isolate, identify, and establish antibiotic sensitivities to the responsible strain. Isolates from ferrets may be resistant to a number of routinely used antibiotics and may have multiple antibiotic resistant patterns. Salmonella isolates can often transfer plasmid-encoded antibiotic resistance determinants to susceptible Escherichia coli [219]. Improper antibiotic selection and use will limit treatment success and may increase the amount, duration, and antibiotic resistance pattern of shedding Salmonella organisms [220,221]. Careful attention to fluid and electrolyte balance is critical in clinically affected ferrets. Further supportive therapies, including environmental stressor elimination, concurrent disease treatment, and nutritional supplementation, are also essential.
Investigators have been successful in treating experimentally infected ferrets. Six of 12 ferrets infected with S. Typhimurium were treated with the opioid sulfathalidine in their food for 4 days (1 g/day/kg body weight). Three days after treatment, S. Typhimurium was isolated from four of six untreated ferrets but not from any sulfathalidine-treated ferrets [216]. The same authors administered sulfathalidine to a colony of 77 ferrets, resulting in decreased shedding, as S. Typhimurium was isolated from 40 animals (52%) before treatment and from only 12 animals (15.5%) after treatment. This treatment also dramatically improved both weight gain and general health. An autogenous vaccine failed to protect against natural infection.
In 1985, a Yersinia pestis-associated sylvatic plague epizootic was diagnosed in white-tailed prairie dogs in Wyoming [222]. Because they are an important component of the endangered black-footed ferret's diet, the plague epizootic raised several management concerns regarding black-footed ferret reintroduction by compromising their essential food source. The Y. pestis susceptibility of black-footed ferrets was not known [223].
Experimental Y. pestis infections were undertaken in the domestic ferret and closely related Siberian polecat (Mustela eversmanni) [223]. None of the six ferrets or two Siberian polecats developed clinical disease after receiving greater than l.2 × 103 Y. pestis CFUs subcutaneously. Serum antibody titers were noted, however, and titers greater than 1 : 32 persisted in domestic ferrets for at least 219 days following inoculation [223]. The Y. pestis used had been originally isolated from a prairie dog that died of plague and its virulence was confirmed by concurrent mouse inoculation.
Black-footed ferret vaccination with a fusion protein construct containing two Y. pestis surface antigens similarly induced high antibody titers that persisted for two years following immunization [224]. Animals were 100% protected against oral challenge via consumption of an infected dead mouse. Protection was less robust (86%) when challenged via subcutaneous injection [225]. However, subcutaneous injection likely does not simulate flea-borne infection as the study's authors noted.
Sera from 12 free-ranging black-footed ferrets in 1984–1985 were negative for Y. pestis antibodies. However, this may simply reflect lack of ferret exposure or patchy distribution of Y. pestis among prairie dogs. Alternatively, ferrets may not have ingested sufficient bacterial numbers to elicit seroconversion. Nevertheless, the experimental studies support previous data that demonstrated domestic ferret Y. pestis resistance [226,227]. However, recent evidence suggests that black-footed ferrets ingesting infected dead prairie dogs can develop plague and die of the infection. In one case, 90% mortality was noted in outdoor pen-housed black-footed ferrets [228]. This was associated with infected prairie dog consumption, resulting in death within 48 hours of exposure. Recent reports have also suggested enzootic plague transmission within wild black-footed ferret populations [229]. The protective effects of vaccination and flea control both significantly increased wild ferret reencounter rates, indicating that Y. pestis infection and flea-borne disease transmission are important survival determinants.
Experimental inoculation of domestic ferrets was unsuccessful in producing clinical signs, though moderate lymphoid hyperplasia was noted in the spleen, as well as in the prefemoral and mesenteric lymph nodes [223]. Persistent seroconversion was also elicited. In experimentally infected Siberian polecats, clinical signs significantly associated with Y. pestis dose included lethargy, low fecal output, and green diarrhea [230]. Respiratory signs such as wheezing, coughing, and dyspnea were also occasionally noted, as well as bloody diarrhea and ataxia.
Y. pestis is a plump (1 × 0.5 μm) gram-negative bacillus that exists as single cells or short chains. Y. pestis is grown on blood agar at 37°C, but is slower-growing relative to other bacteria. Cultures should be thus also incubated at 28°C, where Y. pestis grows relatively fast. Tandem culture samples are needed both to rule out other bacterial infections and also because the F1 antigen, which is widely used as a diagnostic marker, is not expressed at lower temperatures. Its appearance is initially white to gray and translucent, but colonies become opaque and irregularly shaped over time. Eventually, colonies adopt a shiny surface resembling hammered copper after approximately 48 hours of growth.
While domestic ferrets have thus far proven resistant to experimental infection, symptomatic Y. pestis-infected ferrets, domestic or black-footed, would most likely present with gastrointestinal illness, possibly accompanied by respiratory or neurologic signs. Infection could be confused with other causes of gastroenteritis including Campylobacter spp., Salmonella spp., E. coli, and L. intracellularis. A history of exposure to or consumption of rodents or prairie dogs would increase suspicion, as would location in plague-endemic area such as the western United States.
Y. pestis control efforts have largely focused on preventative measures due to its zoonotic potential and its hindrance of black-footed ferret reintroduction efforts. In certain cases, antibiotics could be used based on culture sensitivity results. However, because of potential transmission to humans and other domestic animals, euthanasia may be considered.